This application is based on Application No. HEI 10-221296 filed in Japan, the content of which is hereby incorporated by reference.
1. Field of the Invention
The present invention relates to a halftone dot image discrimination method and image processing device applicable to digital copiers and the like.
2. Description of the Related Art
In digital copiers and the like, image data obtained by reading a document are subjected to various image processing in accordance with the type of document image. Types of document images include, for example, text images, variable density images, halftone dot images and the like, and in order to discriminate these image types, the document image is divided into small block areas. For example, to discriminate whether or not an image is a halftone dot image, the halftone dot image discrimination must be performed on each block area.
Conventionally, the methods used to discriminate a halftone dot image determine whether or not the number of white or black isolated points within a block area exceeds a threshold value.
FIG. 21 illustrates the halftone dot image discrimination method.
In FIG. 21, the block area BE has, for example, a vertical by horizontal size of 9xc3x9741 dots, and the number of isolated points within the block area BE is detected using a 5xc3x975 size isolated dot detection filter FD. In the detection of isolated points, the isolated point detection filter FD is applied to all pixels PX sequentially from the upper left pixel of the block area BE. The pixels matching the pixels in the center of the isolated dot detection filter FD are designated as the target pixels PXT, and these target pixels PXT are checked to determine whether or not they satisfy predetermined conditions.
The predetermined conditions satisfied in this instance determine if a certain target pixel PXT is a white isolated point by determining whether the target pixel PXT has a value larger than or equal to the values of the eight peripheral pixels PX, and whether the value of the center target pixel PXT is greater than or equal to the average value of two pixels PX arrayed in a row along the eight lateral, vertical, and inclined directions. These satisfied conditions also determine if target pixel is a black isolated point by determining whether the target pixel PXT has a value less than or equal to the values of the eight peripheral pixels PX, and whether the value of the center target pixel PXT is less than or equal to the average value of two pixels PX arrayed in a row along the eight lateral, vertical, and inclined directions.
The isolated point detection filter FD must be applied at one time for all image data of the five lines corresponding to the area size. Accordingly, the at least a 4-line memory is used to delay the image data of each line.
In this way the number of determined isolated points is counted. When this count value exceeds a pre-set threshold value, the block area BE is discriminated as a halftone dot area.
In the previously mentioned conventional halftone dot image discrimination method, a 5xc3x975 size filter is used as the isolated dot detection filter FD. The size of the isolated dot detection filter FD is set so as to effectively detect isolated points via the isolated point detection filter FD when the image resolution is 400 dpi.
That is, as shown in FIG. 22(A), when the image resolution is 400 dpi, even a large isolated point SP can be contained within the range of the isolated point detection filter FDa. Accordingly, the isolated point SP can be detected. However, when the image resolution is 600 dpi, as shown in FIG. 22(B), the size of the isolated point detection filter FDb is relatively small.
Therefore, even the identical isolated point SP cannot be contained within the isolated point detection filter FD. as a result, this point cannot be detected as an isolated point SP.
For example, when the size of the isolated point SP is the size shown in FIG. 22(B), the number of isolated points SP becomes [0] regardless of the presence of several isolated points SP, and since the number of isolated points SP is below the threshold value, there is concern that the area will not be discriminated as a halftone dot area. In this case, a discrimination error occurs.
In order to eliminate this problem, the algorithm for detecting the isolated point SP is modified to correct the detection of the isolated point SP even when the isolated point SP is large. In this instance, however, it is necessary to use a circuit or a method for detecting isolated points using a different algorithm at different resolutions. As a result, this alone complicates the algorithms and circuit construction, and is disadvantageous from a cost perspective.
Another methods considered for eliminating this problem increases the size of the isolated point detection filter FD when the resolution is higher. In this instance, however, the line memory must be increased in accordance with the size of the isolated point detection filter FD, circuit construction becomes more complex, and costs rise.
In recent years, image resolution has been generally increasing, and the types of images has tended to increase, such that the aforesaid methods of eliminating the problem have become increasingly disadvantageous from a cost perspective.
An object of the present invention is to eliminate the previously described disadvantages.
In view of the previously described problems, another object of the present invention is to provide a halftone dot image discrimination method and image processing device capable of accurately discriminating halftone dot images at various resolutions even when the resolution increases and when the types of resolutions increase, and which is capable of suppressing cost increases as much as possible.
These and other objects are attained by an image processing device having, an isolated point detection filter for detecting isolated points in image data, and a modifier for reducing the size of the isolated points included in the image data so as to be smaller than the size of the isolated point detection filter.
The objects of the present invention are attained by an image processing device having, an isolated point detection filter for detecting isolated points in image data, a resolution detection unit for detecting the resolution of the image data, and a modifier for reducing the size of the isolated points included in the image data in accordance with the resolution of the image data.